A case-control study is being proposed to evaluate the role of hormonal factors in the etiologies of endometrial and ovarian cancer. Women with cancer will be identified through population-based cancer registries in the states of Washington and Utah. By means of area sampling techniques, control subjects will be identified from the whole of the female population from which the cancer cases arose. Cases and controls will be interviewed with particular attention to estrogen hormone use and reproductive and menstrual history. An attempt will be made to assemble all pathologic material for review and classification by study pathologists.